or Connect
AVS › AVS Forum › Display Devices › Ultra Hi-End HT Gear ($20,000+) › D-cinema Equipment and Theaters › Dolby Atmos Theatre System
New Posts  All Forums:Forum Nav:

Dolby Atmos Theatre System - Page 11

post #301 of 918
Quote:
Originally Posted by Roger Dressler View Post

There's no way to know if the speakers are being driven by channels or objects. Only the mixer knows.

Well, Marc obviously heard a difference.

One of the problems with beds is that there's no standardized way of delivering them. Could be 5.1, 7.1 or higher. Why install 5 front speakers if an unknown number of mixes make use of only 3 of them?

From the Dolby white paper: "Dolby Atmos therefore supports “beds” in addition to audio objects. Beds are effectively channel-based submixes or stems. These can be delivered for final playback (rendering) either individually or combined into a single bed, depending on the
desire of the content creator. These beds can be created in different channel-based configurations such as 5.1, 7.1, or even future formats such as 9.1 (including arrays of overhead loudspeakers)."

With objects a virtually unlimited number of channels is the basis. Better reproduction systems would benefit - for each and every mix.

Anyway, my original point was "realism". Beds might be enough for most mixes but if increased realism is the goal then reverberation has to be presented from more and/or different locations than beds can provide.
Edited by markus767 - 7/4/12 at 4:13pm
post #302 of 918
Quote:
Originally Posted by sdurani View Post

A soundtrack that highlights the new technology. I know that deliberately calling attention to the mix might be an anathema to some mixers, but there are certain soundtracks that make new technologies more obvious than other soundtracks.
Pixar has always been known for tasteful, non-distracting mixes, so nothing about the 'Toy Story 3' soundtrack made it apparent that it was the first 7.1 theatrical mix. By comparison, the first few seconds of 'Transformers 3' (Paramount logo) made it clear that you were hearing separate side and rear channels. Not judging artistic merit, simply pointing out that some soundtracks highlight a new technology more obviously.
Since 'Brave' was a Pixar mix, it is not surprising that some listeners feel that it didn't take "full advantage" of Atmos, based just on its lack of agressiveness. Can't wait to hear what someone like Michael Bay does with Atmos; the results may be gimmicky, but there will be no question that height speakers are in da house.

Sanjay.. I understand the sentiment..it was a bit of a rhetorical comment.

Although I did think the fairly dust during the Disney logo that lept from the sides to the overheards then the other side of the theater was as cool as the 7.1 Paramount logo in TF3.

My point is that Atmos is more than just a large amount of discrete outputs..

5 behind the screen, full range speakers in the Atmo"sphere", bass management, overheads (I noticed a lot more the second time around..), what Dolby refers to as pan throughs, new equalization processors..

All were clearly evident, IMO. smile.gif
Edited by FilmMixer - 7/4/12 at 4:22pm
post #303 of 918
Quote:
Originally Posted by markus767 View Post

Well, Marc obviously heard a difference.

Markus.. I never said I heard a difference.. I said I heard material that was object only.
Quote:
One of the problems with beds is that there's no standardized way of delivering them. Could be 5.1, 7.1 or higher. Why install 5 front speakers if a mix makes use of only 3 of them?

"Brave" did use 5 fronts.. as I suspect will alot of Atmos mixes.
Quote:
From the Dolby white paper: "Dolby Atmos therefore supports “beds” in addition to audio objects. Beds are effectively channel-based submixes or stems. These can be delivered for final playback (rendering) either individually or combined into a single bed, depending on the desire of the content creator. These beds can be created in different channel-based configurations such as 5.1, 7.1, or even future formats such as 9.1 (including arrays of overhead loudspeakers)."

With objects a virtually unlimited number of channels is the basis. Better reproduction systems would benefit.

Anyway, my original point was "realism". Beds might be enough for most mixes but if increased realism is the goal then reverberation has to be presented from more and/or different locations than beds can provide

Understand the point about realism... however, even in the same white paper you quote Dolby recommends beds for reverbs IIRC.

And again, you can easily do object only...

From a work flow standpoint it would make almost no sense to make use of objects for the main dialog track.. and music would also be more difficult to manage.

And I think you would quickly reach the point of rapidly diminishing reutrns trying to get too crazy with multiple object based reverbs... too many pre delays will quickly muddy up the track..
post #304 of 918
Quote:
Originally Posted by FilmMixer View Post

"Brave" did use 5 fronts.. as I suspect will alot of Atmos mixes.

From the "making of" that I saw Brave was a native 7.1 mix which was then remixed and spread in Atmos.

I'm just wondering how it would have sounded if it had been a native top to bottom Atmos mix (which is why I said I felt it didn't take full advantage).
post #305 of 918
Quote:
Originally Posted by sdurani View Post

When I clicked on the on-line schedule for the AMC Van Ness, it lists 'Brave' 3D separately from 'Brave' in Dolby Atmos.
http://www.amctheatres.com/movie-theatres/amc-van-ness-14

Thanks, Sanjay; turns out my second google attempt specifically listed Atmos.

Saw it today; quite impressive though not all that I was expecting after what others have said.

Many times the musical portions of the surround track were less than satisfying due to one or more of: too loud, incipient stress of the speakers, or tonal balance a bit too bright.

The immersiveness was enthralling, and some of the action sounds were really thrilling, like the bears' roaring and the thunder of hoofbeats of Merida's horse.
post #306 of 918
Quote:
Originally Posted by FilmMixer View Post

And I think you would quickly reach the point of rapidly diminishing reutrns trying to get too crazy with multiple object based reverbs... too many pre delays will quickly muddy up the track..

My hope is that guys like you start experimenting with the possiblilities. I'm not aware of reverbs that make use of the additional angles objects can provide. Are there any?
With better speraration the cocktail party effect should help keeping the track transparent. What you describe ("muddy track") is a typical problem of "not enough channels".
Edited by markus767 - 7/5/12 at 3:20am
post #307 of 918
Quote:
Originally Posted by noah katz View Post

The immersiveness was enthralling, and some of the action sounds were really thrilling, like the bears' roaring and the thunder of hoofbeats of Merida's horse.
The main thing I noticed, even more than the height channels, was the pan-through effect (especially with the little blue wisps).
post #308 of 918
Quote:
Originally Posted by tibia View Post

My wife and I just returned from seeing/hearing the ATMOS presentation of "Brave" at the new Cinetopia theatre in Vancouver, WA. The movie was preceded by a well done narrated demonstration of movie sound formats leading up to ATMOS. It was most impressive. Pixar's use of ATMOS in "Brave" is definitely restrained from the demo film so far as extreme pin point source location is concerned. Rather the sound field is the most encompassing i've ever heard without being intrusive.
The showing I attended was not without fault, however. There were two or three occasions where the digital bit stream was momentarily corrupted in some manner resulting in a crackling/static like noise.
I agree with sdurani that it will be real interesting to hear an ATMOS installation where some film director fully exploits it.
Dennis
Perhaps "fully" exploiting it would continually draw attention from the screen ... from the STORY ... and to points around the space?

Jeff
post #309 of 918
Thread Starter 
Quote:
Originally Posted by markus767 View Post

Well, Marc obviously heard a difference.
I should have limited my comment to the front speakers. It will obviously be easy to tell in the arrays whether the sound is a "channel" or an object there.
Quote:
One of the problems with beds is that there's no standardized way of delivering them. Could be 5.1, 7.1 or higher. Why install 5 front speakers if an unknown number of mixes make use of only 3 of them?
The use of beds for certain elements does not preclude the use of pans in the same mix or even at the same time.
Quote:
Anyway, my original point was "realism". Beds might be enough for most mixes but if increased realism is the goal then reverberation has to be presented from more and/or different locations than beds can provide.
That has yet to be shown in a cinema context AFAIK. I would really like to see more examples in this area, though.
post #310 of 918
Quote:
Originally Posted by Roger Dressler View Post

That has yet to be shown in a cinema context AFAIK. I would really like to see more examples in this area, though.

Real life sounds tremendously different from anything I've ever heard in a movie theater - there has to be a reason for that.
post #311 of 918
Thread Starter 
Quote:
Originally Posted by markus767 View Post

Real life sounds tremendously different from anything I've ever heard in a movie theater - there has to be a reason for that.
Probably lots of reasons. The question is, which are they, and how to fix them, assuming they need fixing. Is "real life" what movie making is concerned with reproducing?

Very often, the medium is part of the art form. Oil paintings never look like real life, but that does not diminish their value. Same for 24 fps film. We cannot get music to sound real in a home, so asking it to meet that goal in a cinema may be beyond the brief, particularly as music is an adjunct to the storytelling, not an end in itself. Good question though: what is the definition of success for movie sound reproduction? Used to be "same as the director heard on the dubbing stage." Now it should be something more?
post #312 of 918
Quote:
Originally Posted by Roger Dressler View Post

Probably lots of reasons. The question is, which are they, and how to fix them, assuming they need fixing. Is "real life" what movie making is concerned with reproducing?
Very often, the medium is part of the art form. Oil paintings never look like real life, but that does not diminish their value. Same for 24 fps film.

Very true but why limit ourselves to oil paintings when a new reproduction technique offers oil paintings and much more?
Quote:
Originally Posted by Roger Dressler View Post

We cannot get music to sound real in a home, so asking it to meet that goal in a cinema may be beyond the brief, particularly as music is an adjunct to the storytelling, not an end in itself. Good question though: what is the definition of success for movie sound reproduction? Used to be "same as the director heard on the dubbing stage." Now it should be something more?

I'm more interested in what Atmos could bring to music than what it could bring to movies. As Marc said, there's a disconnect between the little window in the front and the sound being all around the viewer.
post #313 of 918
The more I think about "beds" the less I understand why that feature even exists. Why not simply use "snap to nearest speaker"?

The renderer (on the consumer side) could be made intelligent enough to decide what to do if there's a gross mismatch between object location and actual speaker location. For example, a missing center speaker could be easily transformed to a phantom center by routing the sound to left and right.
Mics for determining speaker locations exist, so the process could be fully automated.
Edited by markus767 - 7/6/12 at 3:08am
post #314 of 918
Thread Starter 
Quote:
Originally Posted by markus767 View Post

The more I think about "beds" the less I understand why that feature even exists. Why not simply use "snap to nearest speaker"?
The bigger distinction between beds and objects can be seen for the surround arrays. Beds makes it easy to spread a sound uniformly across the array. Not so easy with an object.

WRT the front speakers, there is no meaningful difference between "snap to nearest speaker" and beds, except in the latter the mixer specifies the speaker.
Quote:
The renderer (on the consumer side) could be made intelligent enough to decide what to do if there's a gross mismatch between object location and actual speaker location. For example, a missing center speaker could be easily transformed to a phantom center by routing the sound to left and right.
That happens automatically regardless of object or bed/channel. If the intended speaker is not present, the sound is not discarded.
post #315 of 918
Thread Starter 
Quote:
Originally Posted by markus767 View Post

Very true but why limit ourselves to oil paintings when a new reproduction technique offers oil paintings and much more?
We'll have a better answer to that when we see The Hobbit at 48 fps.
Quote:
I'm more interested in what Atmos could bring to music than what it could bring to movies. As Marc said, there's a disconnect between the little window in the front and the sound being all around the viewer.
I'm very much keen to see where music can go with object audio. Should be interesting!
post #316 of 918
Quote:
Originally Posted by Roger Dressler View Post

We'll have a better answer to that when we see The Hobbit at 48 fps.

And rumor has it it will be presented in Atmos as well!
post #317 of 918
Quote:
Originally Posted by Roger Dressler View Post

The bigger distinction between beds and objects can be seen for the surround arrays. Beds makes it easy to spread a sound uniformly across the array. Not so easy with an object.

I believe there's a feature that does just that - play decorrelated versions of an object sound from an array of speakers. So no beds are needed.
Quote:
Originally Posted by Roger Dressler View Post

WRT the front speakers, there is no meaningful difference between "snap to nearest speaker" and beds, except in the latter the mixer specifies the speaker.

So why did they incorporate beds at all?
Quote:
Originally Posted by Roger Dressler View Post

We'll have a better answer to that when we see The Hobbit at 48 fps.

That's not exactly what I meant. Object-based mixes can behave just like an old-fashioned 5.1 or 7.1 channel-based mix but, it can also be much more. A movie in 48fps is 48fps and not 24fps.

By the way, 48fps is probably not enough frames. Now people will say "soap opera effect" but I believe this is just a matter of not enough motion blur and/or depth of field.
Quote:
Originally Posted by Roger Dressler View Post

I'm very much keen to see where music can go with object audio. Should be interesting!

Indeed, but is Dolby actively approaching the music industry?
post #318 of 918
Markus..

You're missing the most obvioous reaons why beds are necessary.

Film sound tracks are obviously made up of thousand of elements per reel (lets say 20 minutes on average.)

When mixing, you need processing, external gear, group panning, etc..

As you start dealing with objects, you need to start concerning yourself with managing resources such as how many objects can a given system handle, etc..

If most of my dialog comes out of the center speaker, and I have 32 channels of dialog premixes, why do I need to tie up ovject resources to do that.

It's true than I can just buss them together and manage that output as a single object,but it's the last thing I want to think about.

In addition, without any kind of menaingful tools (i.e. reverbs with greater than 5 outputs, etc) it's nice to be able to use a standard 7.1 bed to spred that into the array without having to re-invent the wheel.

Not sure why you're so hung up on this... beds are great to have from a work flow stnadpoint, and you're not forced to use them if you didn't want to..

Weve got both options.

The other thing we have to deal with, which is something that still needs to be worked out, is all of the other deliverables we have to create for the stuidos when finishing a film..

Minus mixes, M and E's, crash downs, etc... this is going to be a challenge when we start mixing in Atmos/MDA from the start (instead of going the other way)...
post #319 of 918
Thread Starter 
Quote:
Originally Posted by markus767 View Post

I believe there's a feature that does just that - play decorrelated versions of an object sound from an array of speakers. So no beds are needed.
Yes, MDA has a means to apply decorrelation for each speaker feed. But if the source is an object, there's only one signal driving one speaker, so it still sounds like a point source. It does not spread through the array unless some other function facilitates that.

Here are two ways to do that: MDA can deliver the sound as a channel (bed) which maps to the designated array, or the mixer can apply divergence which spreads the subtended angle of the sound to any chosen amount. They will sound exactly the same if the divergence angle aligns with the array, but divergence is not limited to array boundaries.
Quote:
So why did they incorporate beds at all?
I'm saying that when it comes to the front speakers, beds and "snap to nearest speaker" are just two names for the same function. If you agree that snap is a useful tool, then you agree that beds are useful.
Quote:
Indeed, but is Dolby actively approaching the music industry?
I have seen no mention of it. I think they are well occupied with film for now. However, SRS has demonstrated MDA using music.
Edited by Roger Dressler - 7/8/12 at 1:18pm
post #320 of 918
Thread Starter 
Quote:
Originally Posted by FilmMixer View Post

If most of my dialog comes out of the center speaker, and I have 32 channels of dialog premixes, why do I need to tie up ovject resources to do that.

It's true that I can just buss them together and manage that output as a single object, but it's the last thing I want to think about.
That is something you ought never have to think about, since it is up to the delivery system to handle it automatically in order to optimize the payload.
Quote:
The other thing we have to deal with, which is something that still needs to be worked out, is all of the other deliverables we have to create for the stuidos when finishing a film..
Minus mixes, M and E's, crash downs, etc... this is going to be a challenge when we start mixing in Atmos/MDA from the start (instead of going the other way)...
I know about minus mixes and M and E's, but what are crash downs?
post #321 of 918
Quote:
Originally Posted by FilmMixer View Post

beds are great to have from a work flow stnadpoint

I agree but maybe it's time to rethink workflows in order to advance the art.
post #322 of 918
Quote:
Originally Posted by Roger Dressler View Post

If you agree that snap is a useful tool, then you agree that beds are useful.

No. You can't use multichannel reverbs (and I'm not talking about only 5 channels) with beds the same way you could use them with objects.
post #323 of 918
Quote:
Originally Posted by Roger Dressler View Post

That is something you ought never have to think about, since it is up to the delivery system to handle it automatically in order to optimize the payload.
I know about minus mixes and M and E's, but what are crash downs?

Roger... I agree we shouldn't have to worry about it.. but since I fundamentally disagree with Markus' disdain for beds, it's still a valid concern when using a bed+object workflow with Atmos.. the delivery system still has finite processing resources.

When using an object panner (MDA, Atmos, etc..) to create postional data still requires consideration of all the other things required to get it to sound as you want it to... alot of times the sound isn't just one "object" or sound file...

With the current tools we have at our disposal, and only one playback system for the cinema, object management is still something that needs to be considered.... until I do a project from scratch in the format, however, we can't know if it's something to be concerned with at all, and the tool set for Atmos is evolving so it might be a concern that has no basis.

Crash downs are things like stereo DME's, where you deliver a mono DX track and LoRo MX and FX mixdown... for trailers, archieve, tv spots, internet, etc.

If you are feeding all objects into a renderer there is no seperation like when you create discrete stems... don't know how we would handle this from our organizaitonal work flow at this time.. at this point, you feed the beds and objects into an RMU on the stage to monitor... the simple soulution might be to create 5.1 or 7.1 mixdowns after the film is locked by taking the source workstations off line for each discipline (i.e. only run dialog untis and print the rendered down mix.. two issues with that however are that there isn't alwasy a concrete seperation of elemetns (i.e. some dialog elements end up in the M and E) and it's labor intensive, (i.e. expensive to run that many passes..)

I'm sure Dolby and the mixing community are working on workflow suggestions... I know a white paper with suggestions is upcoming.
post #324 of 918
Quote:
Originally Posted by markus767 View Post

I agree but maybe it's time to rethink workflows in order to advance the art.

You are the one that thinks beds are holding the art back..

I respectfully disagree. smile.gif
post #325 of 918
techradar.AV article, "Is Dolby Atmos the future of cinema sound?" (link), with (just) a few new thoughts about theatrical|home 3D sound systems.
post #326 of 918
Our Atmos or SRS MDA domestic processors aren't available yet so this is a very subjective question.....

Does anyone have any thoughts on what might constitute an optimal speaker layout for heights in a home cinema environment? Would it be 3 fronts and 2 rears....or 2 fronts, 2 sides and 2 rears...or..? Have any companies done work on this and reported their findings?

I'm sure that several of you have opinions on how many height speakers are likely 'optimal', what their best locations would be and the reasons why. I'd be interested to hear some thoughts on this.

In mentioning heights, I'm not talking about over heads, but rather speakers placed high up in the walls near the ceilng.

As regards over heads, I wonder whether many of us are likely to put these into our listening environments? That would need to be done very carefully so-as to avoid hot spotting. I for one, would prefer to go with heights rather than over heads but I'm not sure whether heights would/could perform the same function. Most likely they could given that our listening environment is smaller than a commercial cinema.
post #327 of 918
Quote:
Originally Posted by Kwikas View Post

Have any companies done work on this and reported their findings?
Sean Olive at Harman has done some testing to find out what height speaker locations were prefered by listeners. They ended up with 5 height speakers above a standard 7.1 layout. The centre height goes above the centre speaker. The front heights split the angle between your front speakers and side speakers. The back heights split the angle between your side speakers and rear speakers. He mentioned that centre height speaker was helpful some of the time, though not as much as the others. So I would at least do a grid of 4 height speakers, at the angles mentioned. They can be high up on the walls (at/near ceiling height) or on the ceiling itself and moved in from room boundries (preferred). Either way, the grid will create phantom imaging between them (appear to come from above you) while avoiding the kind of hot-spotting you would get from a speaker directly above you. Increasing the levels of one or two height speakers will allow the overhead sound to be steered in any direction (left, right, front, back, diagonally), based on what's happening in the 7 main speakers.
post #328 of 918
Quote:
Originally Posted by sdurani View Post

Sean Olive at Harman has done some testing to find out what height speaker locations were prefered by listeners. They ended up with 5 height speakers above a standard 7.1 layout. The centre height goes above the centre speaker. The front heights split the angle between your front speakers and side speakers. The back heights split the angle between your side speakers and rear speakers. He mentioned that centre height speaker was helpful some of the time, though not as much as the others. So I would at least do a grid of 4 height speakers, at the angles mentioned. They can be high up on the walls (at/near ceiling height) or on the ceiling itself and moved in from room boundries (preferred). Either way, the grid will create phantom imaging between them (appear to come from above you) while avoiding the kind of hot-spotting you would get from a speaker directly above you. Increasing the levels of one or two height speakers will allow the overhead sound to be steered in any direction (left, right, front, back, diagonally), based on what's happening in the 7 main speakers.

Curious - is there a visual diagram anywhere regarding position?
post #329 of 918
Quote:
Originally Posted by adidino View Post

Curious - is there a visual diagram anywhere regarding position?
Nothing final, just preliminary placement. Last time I checked (early June), Dr. Olive was still doing double blind testing to find final placement. Keep in mind they're using the height extraction in their proprietary surround processing (QuantumLogic), since there is no content with actual height information. Things could change once movies with height channels show up. prelim_height_placement.jpg 14k .jpg file
post #330 of 918
Quote:
Originally Posted by sdurani View Post

Keep in mind they're using the height extraction in their proprietary surround processing (QuantumLogic), since there is no content with actual height information. Things could change once movies with height channels show up.
+1

Of course, it's also possible there may be several competing "height information" channel|speaker configurations at least initially . . . which is unlikely to encourage any speedy consumer adoption of discrete 3D audio! biggrin.gif
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: D-cinema Equipment and Theaters
AVS › AVS Forum › Display Devices › Ultra Hi-End HT Gear ($20,000+) › D-cinema Equipment and Theaters › Dolby Atmos Theatre System